Matrix Metalloproteinase-3 -1171 5A/6A Polymorphism (rs35068180) is Associated with Risk of Periodontitis

Matrix metalloproteinase-3 (MMP3) plays a key role in tissue degradation in periodontitis. The relationship between the MMP3 -1171 5A/6A polymorphism (rs35068180) and periodontitis has been widely studied. However, existing studies have yielded contradictory results. We therefore conducted a meta-analysis to comprehensively investigate these inconclusive findings. Several electronic databases were searched for eligible articles. Seven case-control studies from 6 articles were searched without any language restrictions. Pooled estimates indicated that MMP3 -1171 5A/6A polymorphism is associated with a decreased risk of periodontitis (allelic genetic model: OR = 0.70, 95% CI: 0.62–0.80, Pheterogeneity = 0.315; heterozygous model: OR = 0.50, 95% CI: 0.39–0.65, Pheterogeneity = 0.221; homozygous model: OR = 0.42, 95% CI: 0.25–0.69, Pheterogeneity = 0.265; dominant model: OR = 0.49, 95% CI: 0.38–0.62, Pheterogeneity = 0.238, respectively). Similar results were also found in chronic periodontitis (CP), Asian, Asian&CP, and non-smokers subgroups. Moreover, MMP3 rs35068180 polymorphism might be associated with a lower risk of aggressive periodontitis (AgP) in Asians (allelic genetic model: OR = 0.66, 95% CI: 0.48–0.91, Pheterogeneity = 0.945), and CP in Caucasians and Brazilians. In conclusion, this meta-analysis demonstrates that MMP3 -1171 5A/6A polymorphism may be associated with decreased risk of both CP and AgP in Asians. Large independent studies to replicate these results are necessary to validate these associations in other populations.

gene, located on chromosome 11q22. 2-22.3 (adjacent to the MMP1 and MMP12 genes), is an endopeptidase produced by smooth muscle cells, macrophages and synovial cells 11 . The insertion/deletion of a single adenosine (5A/6A) at position -1171 of the MMP3 promoter region (rs35068180 or rs3025058) can alter its transcription levels 12 . This may affect its ability to degrade connective tissue and lead to the progression of periodontitis. A prioritisation analysis of candidate genes for periodontitis using multiple computational tools identified MMP3 as one of the most promising genes to be involved in periodontitis 13 .
Over the past decade, many genetic association studies have been conducted to evaluate the association between MMP3 -1171 5A/6A polymorphism and the risk of periodontitis. However, the results of these studies were inconsistent. The discrepancies among studies may be due to the relatively small sample size in each investigation, as well as varying population characteristics. Therefore, we performed a meta-analysis to investigate the potential association of MMP3 -1171 5A/6A polymorphism with the risk of periodontitis.

Results
Characteristics of eligible studies. Seven case-control studies from six publications were identified in the meta-analysis (a flow diagram is shown in Fig. 1). Basic characteristics and the quality assessment scores of the selected articles are all listed in Table 1. Overall, all of the included studies were moderateto high-quality, with a mean quality score of 29.1/40 (range from 27 to 31). Of these case-control studies, four studies including 963 cases and 1227 controls were undertaken in Asians [14][15][16][17] , two studies containing 183 cases and 402 controls were conducted in a mixed Brazilian population 18,19 , and one study with 67 cases and 202 controls was performed in a Caucasian population 19 . Three studies enrolled non-smokers, which permitted smoking to be analysed as a risk factor 14,16,18 . Five eligible studies mentioned quality controls of the genetic analyses, such as blind genotyping, validation of genotyping accuracy, and random replications [15][16][17]19 . Two studies supplied only the allele frequency to compute OR in the allelic genetic model 15,19 . The control populations of three studies were not consistent with Hardy-Weinberg equilibrium (HWE) [16][17][18] . The minor allele frequency (MAF) for controls and genotype distributions of analysed polymorphism in different ethnicities are all listed in Table 2.
In addition, when analysis was limited to the three studies that recruited non-smokers, we also found that the -1171 5A/6A polymorphism was associated with a decreased risk of periodontitis in nonsmokers (6A allele vs.

Sensitivity analysis and publication bias. Sensitivity analysis was performed by omitting individ-
ual studies to assess the effect of each publication on the overall results. No single study changed the pooled ORs qualitatively, which suggests that the results of our meta-analysis are accurate.

First author
Year Country Ethnicity Begger's funnel plot and Egger's test were performed to detect the publication bias of included studies. As shown in Fig. 5, the shape of the funnel plot did not reveal any evidence of obvious asymmetry. In addition, the results of Egger's test did not show any evidence of publication bias.     22 . The relationship between MMP3 -1171 5A/6A polymorphism and periodontitis has been widely investigated. However, previous studies have yielded contradictory results. We therefore performed this meta-analysis to comprehensively evaluate these inconclusive findings. To the best of our knowledge, this is the first meta-analysis to have investigated the association between MMP3 -1171 5A/6A polymorphism and the susceptibility to periodontitis. Our meta-analysis was based on seven case-control studies that focused on the link between MMP3 -1171 5A/6A polymorphism and periodontitis risk and included a total of 1,213 cases and 1,831 controls. The results of our meta-analysis revealed that -1171 5A/6A polymorphism is associated with a decreased risk of overall periodontitis; this association was consistently significant in all subgroups in analyses that were stratified by periodontitis type, ethnicity and smoking status.  Although phenotypic differences exist between AgP and CP in terms of the genetic predisposition to periodontitis, some scholars believe that both chronic and aggressive periodontitis may share some susceptibility genes 23 . Furthermore, some genetic polymorphisms (e.g., TNF-α -308G/A and Fcγ receptor IIIB NA1/NA2) have been shown to have similar effects on AgP and CP 24,25 . In this report, we first analysed the effect of MMP3 -1171 5A/6A polymorphism on two forms of periodontitis. We found that this locus was associated with a decreased risk of periodontitis under four genetic models. Additionally, -1171 5A/6A showed a trend towards protective factor for periodontitis in the recessive model, although the finding was not statistically significant (p = 0.06). When the analysis was limited to the four HWE studies, the association between -1171 5A/6A polymorphism and periodontitis was still significant in the allele model, indicating the robustness of this meta-analysis. Several subgroup analyses were subsequently conducted, and the associations were similar in AgP and CP subgroups in the allele comparison. Additional well-designed studies are required to strengthen our understanding of the association between -1171 5A/6A polymorphism and AgP. Although the minor allele frequency (MAF) and genotype distributions in the included studies for -1171 5A/6A varied by ethnicity (Table 2); the stratified analysis by ethnicity showed a significant association with CP in all three ethnicities. However, all the positive results for the Caucasian and Brazilian mixed populations were derived from a single study. Therefore, future studies with larger samples are warranted to confirm these associations in the Caucasian and Brazilian mixed populations. Tobacco smoking is thought to be an important modifier of the pathogenesis of periodontitis 5 . In further stratified analyses, only the non-smoker subgroup was included to eliminate the possible confounding effects of smoking, and similarly significant findings were found.

Form of Disease
The role of MMP3 in the pathogenesis of periodontitis has been widely investigated. Toyman et al. showed that MMP3 levels in the gingival crevicular fluid (GCF) increased in periodontitis cases and played a role in tissue destruction 26 . Recent findings have demonstrated that clinical improvements after non-surgical periodontal therapy are accompanied by a reduction of MMP3 in GCF in both CP and AgP patients 27,28 . Some meta-analyses have investigated the associations between MMP3 -1171 5A/6A polymorphism and systemic diseases [29][30][31][32][33] and have demonstrated that this polymorphism is associated with a decreased risk of gastrointestinal cancer 30 , abdominal aortic aneurysm 31 , myocardial infarction 33 and head and neck cancer 32 . All of these diseases have been shown to be positively associated with periodontitis in several epidemiological studies [34][35][36][37] . Our results suggest an association between the transcriptionally more active 5A allele and periodontitis, substantiating previous functional findings regarding the MMP3 rs35068180 variant. We speculate that rs35068180 may influence the development of periodontitis through a similar molecular mechanism. In vitro assays of promoter activity have shown that the 5A allele has a two-fold higher promoter activity than the 6A allele 38 . Based on these results, our findings are biologically plausible.
The following limitations of the meta-analysis should be recognised. First, we pooled the data based on unadjusted information, while a more precise analysis needs to be conducted if individual data are available. Second, the number of subjects included in this meta-analysis was relatively small for AgP, Caucasian and Brazilian mixed population subgroups. Hence, the results should be interpreted with caution. Third, the genotypic distribution of the controls also deviated from HWE in three of the studies.
In conclusion, this meta-analysis suggests that MMP3 -1171 5A/6A polymorphism contributes to a decreased susceptibility to CP, especially in Asian populations. Furthermore, this polymorphism may decrease the risk of AgP in Asians. Replication studies of our results in large independent populations are necessary to validate this association. Methods Search strategy. Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were followed 39 . A systematic search of studies that addressed the associations between MMP3 -1171 5A/6A polymorphism and periodontitis was performed. The literature was retrieved in electronic bio-medical databases including Google Scholar, PubMed, Embase, Medline and China National Knowledge Infrastructure (the last search update was 19 Apr 2015). A combination of the terms "Matrix Metalloproteinases", "MMP", "periodontal diseases", "periodontitis", and "polymorphism" were entered both as text words and as Medical Subject Heading components. No language restriction was placed on the search. Manual searches for references cited in published review and original publications were also performed.
Inclusion and exclusion criteria. The following two inclusion criteria were used for all retrieved articles: (1) use of a case-control, cohort, nested case-control, or cross-sectional design; and (2) genotype distribution or allele frequency was available from the published data for both cases and controls. The exclusion criteria were as follows: (1) lack of usable allele frequency or genotype distribution data; and (2) lack of control subjects. Data extraction. Information was carefully extracted independently from all eligible publications by two investigators. Disagreements were resolved through discussion among the authors to achieve a consensus. The following characteristics were collected for each study: the first author's family name, publication year, ethnicity (categorised as Caucasian, Asian and mixed), numbers of cases and controls, type of periodontitis (CP or AgP), distribution of genotypes and alleles, smoking status, and genotype identification method.
Study quality assessment. The quality of each study was evaluated independently by two authors according to the modified STROBE quality score systems 40 . Forty assessment items were used for assessing study quality with scores ranging from 0 to 40. Studies were classified as low-(scores of 0-19), moderate-(scores of 20-29) and high-quality (scores of [30][31][32][33][34][35][36][37][38][39][40]. Any disagreement was resolved by discussion among the authors to reach a consensus. The details of the modified STROBE quality score system are listed in Supplementary Table S1. Statistical analysis. Odds ratios (ORs), along with 95% confidence intervals (CIs), were used to assess the strength of the association between the MMP3 gene -1171 5A/6A polymorphism and periodontitis risk. The Z-test was used to determine the statistical significance of the pooled OR. The pooled ORs were performed for the following genetic models: the mutant allele versus the wild type, homozygote contrast, heterozygote contrast, dominant model and recessive model. In addition, the HWE in the healthy control populations was tested within each study. We applied the chi-squared method to assess if the genotype distribution in the control group was in HWE. Subgroup analyses were also performed for the HWE in controls, type of periodontitis, ethnicity and smoking status.
Both the fixed-effects model (the Mantel-Haenszel method) 41 and the random-effects model (the DerSimonian and Laird method) 42 were used to calculate the pooled ORs. For each genetic comparison, the heterogeneity between studies was assessed using the chi-square based Q-test and was quantified by the I 2 statistic. The fixed-effects model was used when heterogeneity was not significant (P heterogeneity > 0.10 or I 2 < 50%); otherwise, the random-effects model was applied. Sensitivity analysis was performed to assess the stability of the association. Begg's funnel plot 43 and Egger's weighted regression method 44 were applied to evaluate the underlying publication bias. All statistical analyses were conducted with STATA software, version 11.0 (StataCorp, College Station, TX, USA), using two-sided p-values.